The author(s) shown below used Federal funding provided by the U.S. Department of Justice to prepare the following resource: Document Title: Advanced Statistical Population Genetics Methods for Forensic DNA Identification Author(s): Noah A. Rosenberg, Ph.D. Document Number: 253932 Date Received: October 2019 Award Number: 2014-DN-BX-K015 This resource has not been published by the U.S. Department of Justice. This resource is being made publically available through the Office of Justice Programs’ National Criminal Justice Reference Service. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice. Basic Research and Development in Forensic Science for Criminal Justice Purposes Department of Justice, Office of Justice Programs National Institute of Justice NIJ SL # SL001082 NIJ‐2014‐3744 Award # 2014‐DN‐BX‐K015 ADVANCED STATISTICAL POPULATION GENETICS METHODS FOR FORENSIC DNA IDENTIFICATION Prepared by: Noah A. Rosenberg, PhD Principal Investigator Department of Biology Stanford University 371 Gilbert Building, Room 109 Stanford, CA 94305‐5020 Tel: 650 721 2599 Email: [email protected] Prepared on: January 10, 2019 Recipient Organization: Board of Trustees of the Leland Stanford Junior University Stanford University 3160 Porter Drive, Suite 100 Palo Alto, CA 94304‐8445 Final Progress Report Project Period: 01/01/2015 – 12/31/2018 Signature of Submitting Official: Robert Loredo, Contract and Grant Officer 01/24/2019 This resource was prepared by the author(s) using Federal funds provided by the U.S. Department of Justice. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice. Major project goals and objectives The project focused on application of population‐genetic methods of population structure analysis and genotype imputation to problems in forensic genetic analysis. The objectives of the work were (1) to determine the population structure information present in the CODIS forensic markers, (2) to examine cross‐compatibility of microsatellite databases with new single‐nucleotide polymorphism (SNP) markers by use of techniques of genotype imputation, and (3) to improve the population‐genetic basis of advanced forensic genetics techniques such as relatedness profiling. Research accomplishments (1) Population structure and CODIS. We performed an extensive analysis of CODIS microsatellite genotypes in a worldwide panel of individuals from the Human Genome Diversity Panel. We focused on comparing the inference of population structure from CODIS microsatellites to similar inferences made on non‐CODIS microsatellite sets of similar size, relying on the data from past microsatellite studies from our group (NA Rosenberg et al. 2002 Science 298:2381‐2385; NA Rosenberg et al. 2005 PLoS Genetics 1:660‐671; NA Rosenberg 2011 Human Biology 83:659‐684; TJ Pemberton et al. 2013 G3: Genes, Genomes, Genetics 3:891‐907). Our comparative microsatellite analysis demonstrated that the CODIS microsatellites contain a nontrivial level of ancestry information, similar to that of random microsatellite sets. More generally, we found that the level of information about individual identity in a marker set is correlated with the amount of ancestry information. These findings were published in a leading high‐impact journal (BFB Algee‐Hewitt et al. 2016 Current Biology 26:935‐942). They received news coverage in a story in Forensic Magazine. 2 This resource was prepared by the author(s) using Federal funds provided by the U.S. Department of Justice. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice. (2) Record matching between CODIS profiles and SNP profiles. For our second major study of the worldwide panel of CODIS microsatellites, we showed that a CODIS genotype profile can be linked to a corresponding profile of SNP genotypes, even though the profiles share no markers in common. This “record matching” proceeds from the fact that correlations between microsatellites and SNPs enable partial imputations of the microsatellites from the SNPs. The imputed information available on microsatellite genotypes from SNP genotypes then accumulates across the microsatellites to permit profile matching between microsatellite and SNP databases. Thus, two databases, one on the CODIS microsatellites and one on SNP genotypes, could potentially be used to identify a SNP profile as belonging to the same individual as the contributor of a microsatellite profile. This work, which enhances potential for design of backward‐compatible new forensic marker systems and raises privacy issues in cross‐database matching problems, was also published in a high‐ impact journal (MD Edge et al. 2017 Proceedings of the National Academy of Sciences USA 114:5671‐ 5676). The study received news coverage, in Forensic Magazine, Pacific Standard, Stanford Report, and the University of Michigan Health Lab Report. It shed light on a previously unknown connection between forensic genetic data and biomedical and genealogical genetic data, a topic that subsequently came to widespread attention in 2018 when a similar type of connection was exploited for use in identifying crime suspects on the basis of publicly available genetic data of their distant relatives. (3) Record matching of relatives between CODIS profiles and SNP profiles. Our third major study extended our analysis of record matching to close relatives. In particular, we demonstrated that a CODIS genotype profile can be connected to a SNP profile of a close relative, even though CODIS and SNP databases share no markers. The study capitalized both on correlations between microsatellites and SNPs and on classic theory of the effect of relatedness on the identity or non‐identity of genotypes in distinct individuals. The study has as a consequence that SNP profiles can potentially be used to perform 3 This resource was prepared by the author(s) using Federal funds provided by the U.S. Department of Justice. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice. relatedness matching calculations against a database of STR profiles, and vice versa, with potential for success in a nontrivial fraction of cases. It also expands the privacy concerns revealed by our initial study of record matching between SNP and STR profiles. We reported this paper in Cell, one of the leading journals in the biological sciences (J Kim et al. 2018 Cell 175: 848‐858). We are pleased that the paper received widespread media coverage, both for its potential to improve forensic searches and for its attention to privacy concerns. News stories appeared in the scientific press, in Nature, New Scientist, Science, Scientific American, Wired, as well as in mainstream news outlets such as CNN, PBS NewsHour, and the Wall Street Journal. Summary of research contributions. We are pleased to have completed major research papers concerning all three of our objectives. Broad attention to the work, as evidenced by its appearance in high‐profile journals (Cell, Current Biology, Proceedings of the National Academy of Sciences USA), and news coverage of all three main papers, illustrates the interest in topics at the intersection of population genetics and forensic genetics and the potential of population‐genetic approaches for advancing forensic genetics. Training and professional development The project afforded an opportunity to train postdoctoral researchers and PhD students at the intersection of population genetics and forensic genetics. Dr. Bridget Algee‐Hewitt, Dr. Jaehee Kim, and Dr. Michael Edge all conducted research on the project. Graduate student Michael Edge has graduated from Stanford with a Ph.D. in Biology, entitled “Pick up the pieces: combining information from multiple genetic loci,” and receiving the Department of 4 This resource was prepared by the author(s) using Federal funds provided by the U.S. Department of Justice. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice. Biology’s Samuel Karlin Prize for an outstanding Ph.D. thesis in mathematical biology. He has advanced to perform postdoctoral work at the University of California, Davis. Postdoctoral researcher Bridget Algee‐Hewitt has completed her postdoctoral training. She has advanced to a position as a Senior Research Scientist in the Center for Comparative Studies in Race and Ethnicity at Stanford University. Postdoctoral researcher Jaehee Kim is continuing in her studies. Publications We have reported three major publications, on population structure inference in the CODIS markers (Algee‐Hewitt et al. 2016), record matching between CODIS and SNP databases (Edge et al. 2017), and record matching of relatives between databases (Kim et al. 2018). A fourth study extends the principles uncovered in our population structure analysis to produce general results concerning the behavior of the population structure statistic FST (Alcala & Rosenberg 2017). BFB Algee‐Hewitt, MD Edge, J Kim, JZ Li, NA Rosenberg (2016). Individual identifiability predicts population identifiability in forensic genetic markers. Current Biology 26:935‐942. MD Edge, BFB Algee‐Hewitt, TJ Pemberton,